riscv32 EXT4 splat, 6.8 regression?

[Nam Cao]

On 2024-04-17 Andreas Dilger wrote:
> On Apr 16, 2024, at 4:36 PM, Nam Cao <namcao at linutronix.de> wrote:
> > However, I am confused about one thing: doesn't this make one page of
> > physical memory inaccessible?
> > 
> > Is it better to solve this by setting max_low_pfn instead? Then at
> > least the page is still accessible as high memory.  
> 
> Is that one page of memory really worthwhile to preserve?  Better to
> have a simple solution that works,

Good point.

> maybe even mapping that page
> read-only so that any code which tries to dereference an ERR_PTR
> address immediately gets a fault?

Not sure about this part: it doesn't really fix the problem, just
changes from subtle crashes into page faults.

Let me send a patch to reserve the page: simple and works for all
architectures.

Best regards,
Nam